Skip to content

Fix tests with cypari 2.2.6 and pari 2.19 - #42977

Merged
vbraun merged 2 commits into
sagemath:developfrom
antonio-rojas:cypari-2.2.6
Oct 18, 2026
Merged

vbraun merged 2 commits into
sagemath:developfrom
antonio-rojas:cypari-2.2.6

Conversation

@antonio-rojas

Copy link
Copy Markdown
Member

cypari 2.2.6 made some changes in precision handling that break some tests due to numerical noise. Add some tolerance to fix them.

cypari 2.2.6 made some changes in precision handling that break some tests due to numerical noise. Add some tolerance to fix them.
@github-actions

github-actions Bot commented Oct 14, 2026 •

Copy link
Copy Markdown

Documentation preview for this PR (built with commit b32ecf0; changes) is ready! 🎉
This preview will update shortly after each push to this PR.

@dimpase

dimpase commented Oct 15, 2026

Copy link
Copy Markdown
Member

OK, we can use this in #42960, instead of .... While you are at it, there are # long time tags these based on testing done ~15 years ago, can these tags be removed now?

@antonio-rojas

Copy link
Copy Markdown
Member Author

OK, we can use this in #42960, instead of ....

I'm not sure I understand what you're agreeing to. There are 3 options:

While you are at it, there are # long time tags these based on testing done ~15 years ago, can these tags be removed now?

They are definitely not # long time on my machine, I don't know about CI.

@dimpase

dimpase commented Oct 15, 2026

Copy link
Copy Markdown
Member

Can we go with the 1st option you propose here?

@dimpase

dimpase commented Oct 15, 2026

Copy link
Copy Markdown
Member

regarding # long time tags, something that took a second or 2 in 2011 is probably 10 times faster now, so the tag can go. It's up to you, how to deal with it, it's just a suggestion

Use common leading digits for PARI output across cypari2 versions and mark implementation-dependent L-function cost estimates as random. Check precision preservation, sort unordered singular values, and verify generator norms independently of PARI's choice of representatives.
@antonio-rojas

Copy link
Copy Markdown
Member Author

Can we go with the 1st option you propose here?

done

@dimpase

dimpase commented Oct 15, 2026

Copy link
Copy Markdown
Member

OK, fine. I'll rebase #42960 on this one, and proceed from there, checking what else, if anything, should be dropped there

vbraun pushed a commit to vbraun/sage that referenced this pull request Oct 16, 2026
cypari 2.2.6 made some changes in precision handling that break some
tests due to numerical noise. Add some tolerance to fix them.

URL: sagemath#42977
Reported by: Antonio Rojas
Reviewer(s): Dima Pasechnik
@vbraun
vbraun merged commit 38986bc into sagemath:develop Oct 18, 2026
25 checks passed
@antonio-rojas
antonio-rojas deleted the cypari-2.2.6 branch October 18, 2026 16:24
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants